From: Axel Kohlmeyer (
Date: Thu Sep 11 2014 - 08:38:34 CDT

On Tue, Sep 9, 2014 at 7:18 PM, Dominik Bottländer
<> wrote:
> Hello Axel,
> I will be glad to try out the alternative LAMMPS output file format and to
> render support in improving its native import into VMD. Please provide me
> with the necessary information on how to use the alternative file format and
> let me know what specific checks you want me to perform.

it will take me a little while and most likely also some changes to
LAMMPS may be required.
i will get back to you off-list, when i have something ready to try.


> With kind regards,
> Dominik
> Am 08.09.2014 19:45, schrieb Axel Kohlmeyer:
>> dominik,
>> On Mon, Sep 8, 2014 at 1:07 PM, Dominik Bottländer
>> <> wrote:
>>> Hello John,
>>> thank you for getting back to me about the "physical_time" field of the
>>> molinfo command! I will definitely use the field to store the physical
>>> instants of time of my simulations and to access them in my trajectory
>>> analysis tcl-scripts.
>> since you are using LAMMPS, would you be willing to change to a
>> different file format for your work?
>> and make some tests and help with improving this format to import this
>> information natively into VMD?
>> LAMMPS supports a variety of formats (either directly or through an
>> interface to VMD's molfile plugin library. i would have to discuss
>> with other developers, whether an extension to the native file formats
>> can be done or whether it would only be possible through the molfile
>> plugins. this discussion reminded me of some inquiries on the LAMMPS
>> mailing list, where users that use variable time steps, were looking
>> for a way to have an (approximately) even spacing between frames in
>> time and not time steps.
>> so there is an opportunity to add some real value to both LAMMPS and VMD.
>> axel.
>>> Due to your hint I noticed that I had included a link to the VMD 1.7
>>> User's
>>> Guide in my original post on this topic. That must indeed have been a
>>> shock!
>>> :-) The link was my first search result when I quickly looked up "vmd
>>> molinfo" on Google. I use VMD 1.9.1 and usually also consult the
>>> pdf-version
>>> of the respective current User's Guide.
>>> Have a great day!
>>> Dominik
>>> Am 08.09.2014 17:08, schrieb John Stone:
>>>> Dominik,
>>>> Yes, the "physical_time" field is supported by the molinfo command.
>>>> You will do better to refer to the current VMD documentation rather
>>>> than the very old VMD 1.7 user's guide that you had linked to below.
>>>> If you're using VMD 1.7 (that would be a shock), then you had best
>>>> upgrade to a current version! :-)
>>>> Cheers,
>>>> John Stone
>>>> On Sat, Sep 06, 2014 at 08:36:43AM +0200, Dominik Bottländer wrote:
>>>>> Dear Axel, Dear John, Dear Eddi,
>>>>> Thank you so much for your quick reply to my request and
>>>>> the helpful information you provided!
>>>>> @ Eduard Schreiner: You advised me of the true reason for
>>>>> the missing physical time information in VMD by confirming
>>>>> my assumption that the DCD file format does not contain
>>>>> the physical instants of time of the simulation.
>>>>> @ Axel Kohlmeyer: I actually wrote my own tcl-script to
>>>>> calculate the RMSD at the different frames of the
>>>>> respective trajectory to get some experience with
>>>>> tcl-programming in VMD and to specifically implement the
>>>>> position based RMSD definition in accordance with the
>>>>> GROMACS tool "gmx rms"
>>>>> (
>>>>> @ John Stone: Is "physical_time" a prescribed additional
>>>>> keyword of the "molinfo"-command
>>>>> (
>>>>> referring to the physical time field or is it possible to
>>>>> specify additional keywords of arbitrary names for the
>>>>> "molinfo"-command in order to store the physical time
>>>>> field or other data?
>>>>> Thank you very much for your answers in advance!
>>>>> With kind regards,
>>>>> Dominik Bottländer

Dr. Axel Kohlmeyer
College of Science & Technology, Temple University, Philadelphia PA, USA
International Centre for Theoretical Physics, Trieste. Italy.